Alcohol Estimates 2011

8/4/2019 Alcohol Estimates 2011

1/68

Topography of Drinking Behaviours in England


2/68

1

Topography of Drinking Behaviours in England

Synthetic estimates of numbers and proportions of abstainers, lower risk,increasing risk and higher risk drinkers in local authorities in England

August 2011

Executive Summary

Previous research has demonstrated that a sizable proportion of England s populationdrink alcohol at a level associated with a range of health, crime and economicoutcomes. Men who regularly drink over 50 units per week (or eight units per day) andwomen who regularly drink over 35 units per week (or six units per day) are most atrisk of developing alcohol-related illness or injuries or being admitted to hospital.Levels of alcohol consumption (and associated harms) have changed over the lastdecade and so updated estimates are necessary to understand the current situation.

This report presents updated model-based figures (known as synthetic estimates) forthe numbers and proportions of abstainers, lower risk (sensible), increasing risk(hazardous) and higher risk (harmful) drinkers for all local authorities in England for2008 (see Table 1 for definitions of abstainers and different drinking levels). Theseestimates, commissioned by the Department of Health and produced by the NorthWest Public Health Observatory, replace the harmful and hazardous drinking estimatesbased upon data for 2000/02.

This experimental data is intended to help Directors of Public Health in local authoritiesunderstand the levels of alcohol use and misuse in their areas, and support thedevelopment and monitoring of the effectiveness of local and national alcoholstrategies to tackle the harms caused by alcohol. The estimates should be used inconjunction with local intelligence about alcohol use.

These latest estimates of reported drinking behaviour use multinomial regressionmodelling procedures to generate local authority level data. Data in the model includethe General Lifestyle Survey (2008), levels of hospital admissions resulting fromalcohol consumption, population demographics (age, sex, ethnicity), levels of deathsrelating to alcohol and levels of deprivation.

The results presented in this report show that there are large differences in estimatesof reported drinking behaviour between local authorities. Using self reported data onalcohol consumption, we calculate that:

o 6.3% of people in the Isles of Scilly and 6.4% of people in Mid Devon abstaincompared to 48.0% in Newham and 35.4% in Brent.

o 37.5% of people in Newham and 43.6% of people in Tower Hamlets are lower


3/68

2

risk drinkers compared to 74.3% in Tendring and 74.1% in Waveney.1o 8.6% of people in Brent and 8.8% of people in Slough are increasing risk

drinkers compared to 30.9% in Exeter and 30.8% in the Isles of Scilly.o 2.0% of people in Peterborough and 2.2% of people in both Weymouth and

Portland and Kings Lynn and West Norfolk are higher risk drinkers compared to11.7% in both Harrow and Hounslow.

The results presented in this report also show that the picture is very different whenfocusing on the drinking population alone in each local authority (i.e. without includingthose who are abstainers). For example, it is estimated that 16.4% of people inRedbridge are increasing risk drinkers. When limited to the drinking population only, itis estimated that 24.5% of drinkers in Redbridge are increasing risk drinkers. Thecorresponding figures for higher risk drinkers in Redbridge are 6.7% and 10.0%. Littleis known about the characteristics of those who abstain, why they choose to do so andwhat proportion were previously alcohol consumers.

Alcohol consumption is mediated by a range of factors which give rise to the

differences in the percentage of people who abstain and who drink at lower, increasingand higher risk levers both across local authorities and over time. A considerable driverof these differences is a change in wine consumption in terms of increases in theamount consumed generally, increases in the alcohol by volume of wine2 and becausethe biggest consumers of wine are the more affluent subgroups of the population.

The effectiveness of actions to tackle alcohol locally and nationally need to considerhow they impact on drinkers behaviour. Evaluations must distinguish betweenchanges in levels of abstainers which may simply result from demographic changes inage and ethnicity, and interventions which are genuinely moving higher risk andincreasing risk drinkers into a lower risk or abstainer category. Furthermore, in the

current economic climate, it is possible that changes in the level of drinking are relatedto changes in income and employment status.

Levels of abstinence have increased from 10% in 1998 to 15% in 2009 (as measuredby the General Household Survey/General Lifestyle Survey). Alcohol abstainers makeup a significant minority in many localities though they are often ignored in discussionsabout alcohol. The impact of alcohol abstainers must be considered in order to ensurethat policies do not segregate major parts of some communities.

1Small proportions of lower risk drinkers in a local authority can result from higher levels of increasing and

higher risk drinkers but can also result from high levels of abstainers in a locality.2

Alcohol by volume: standard measure of how much alcohol is contained within an alcoholic drink.


4/68

3

1. Introduction

The recently published White Paper, Healthy Lives, Healthy People,3sets out the strategy forpublic health in England and highlights the significant impact of alcohol on society. Accordingto the National Audit Office, over 10 million adults in England drink more alcohol than therecommended daily limit, with the House of Commons Health Committee reporting that 2.6

million of them are drinking more than twice this.

4

The risks of alcohol misuse are wellestablished and place a huge burden on the NHS at an estimated cost of around 2.7 billion ayear.5 The number of hospital admissions attributable to alcohol misuse was 1.1 million in2009/10, a 100% increase since 2002/03.6 Alcohol misuse also contributes to 1.2 millionincidents of violent crime a year, 40% of domestic violence cases and 6% of all roadcasualties.7

National indicators for alcohol are expected to be a part of the Public Health OutcomesFramework for England. 8 Furthermore, as Directors of Public Health move into localauthorities they will take responsibility for commissioning alcohol misuse prevention andtreatment services in collaboration with local commissioning consortia. Such commissioningrequires data on the extent and characteristics of the populations alcohol consumption ineach locality. Currently, there are no direct measures of how many people in local areas aredrinking alcohol or are drinking above the recommended limits.9 However, national surveydata, demographic information and information on alcohol-related mortality and morbidity canbe used in combination to generate estimates of the number of people using different levels ofalcohol in each local authority. The synthetic estimates of drinking behaviour reported hereare therefore intended to help inform the development, targeting and monitoring of theeffectiveness of local and national alcohol strategies.

This report presents new synthetic estimates of increasing risk drinkers (previously referred toas hazardous drinkers) and higher risk drinkers (previously referred to as harmful drinkers) atlocal authority level in England and, for the first time, also includes estimates of the number ofabstainers and lower risk drinkers in each local authority. The estimates were commissionedby the Department of Health and produced by the North West Public Health Observatory.Currently the Department of Health recognises four increasing levels of alcohol consumptionand definitions of these are given in Table 1, along with the operational definitions used inorder to create these synthetic estimates. A further category of drinker (although not exclusiveof those detailed above) is dependent drinkers. The feasibility of producing estimates of levelsof alcohol dependence by local authority, derived largely from models using the AdultPsychiatry Morbidity Survey (2007) is currently being considered by the North West PublicHealth Observatory, with alcohol dependency defined as a self reported AUDIT (Alcohol UseIdentification Test) score of 10 or more and a score of four or more using the SADQC

3Healthy Lives, Healthy People. Department of Health. November, 2010.

4Making alcohol a health priority. Opportunities to reduce alcohol harms and rising costs. Alcohol Concern.

January, 2011.5

Investing in Alcohol Treatments: reducing costs and saving lives. Alcohol Concern, 2010.6

Alcohol-Related Hospital Admissions. North West Public Health Observatory and Centre for Public Health,2010.7

Alcohol: First Report Session of 2009-10. Stationery Office Limited, 2009.8

Healthy Lives, Healthy People: Transparency in Outcomes. Department of Health. December, 2010.9

Units and you. Department of Health, 2008.


5/68

4

(Severity of Alcohol Dependence Questionnaire, community version).10 The Department ofHealth currently uses the ANARP (Alcohol Needs Assessment Research Project)methodology to estimate the potentially treatable population, a methodology which uses a cutoff score on AUDIT of 16 to identify moderately and severely dependent drinkers.11

Table 1: Alcohol drinking categorisation and definitionsDrinkingcategory

Government definitions Operational definitions

Abstainers No Government definition forabstinence exists.

A person whose weekly alcohol consumption wasreported in the General Lifestyle Survey as 0 units overthe previous 12 months.

Lower risk Men who regularly drink no morethan 3 to 4 units per day and womenwho regularly drink no more than 2 to3 units per day.* Weekly limits are nomore than 21 units per week for aman and 14 units per week for awoman .**

A man whose average weekly alcohol consumption wasreported in the General Lifestyle Survey as >0 and 0 and21units to 14units to 50 units inthe previous 12 months.

A woman whose average weekly alcohol consumptionwas reported in the General Lifestyle Survey as >35units in the previous 12 months.

*Details of definitions based on units of alcohol are in: Safe, Sensible, Social Consultation on further action.Department of Health, 2008.**Personal communication. Department of Health. Although weekly lower risk drinking limits are recognised bythe Department of Health they no longer publicise them through any media.

2. Methodology

A methodological summary is presented in Appendix 1, while details of the modelling andcalculation of confidence intervals are presented in Appendices 2 and 3 respectively.

10Adult Psychiatric Morbidity in England, 2007. Results of a household survey. NHS Information Centre, 2009.

11Alcohol Needs Assessment Research Project. Department of Health, 2005.


6/68

5

3. Results and Discussion

Synthetic estimates for the number of abstainers, lower risk drinkers, increasing risk drinkersand higher risk drinkers are presented in Tables 2 to 10 for each local authority in England bynine geographical areas: East Midlands, East of England, London, North East, North West,South East, South West, West Midlands and Yorkshire and Humber. More specifically, these

tables provide for every local authority: 1) the estimated percentage of the population that fallwithin each of the four drinking categories; 2) the estimated number of people who fall withineach of the four drinking categories; and 3) for drinkers only, the percentage of the populationthat fall within each of the three drinking categories. These percentages with theircorresponding 95% confidence intervals are detailed in Appendix 4.

The results show large differences in estimates of drinking behaviour between localauthorities. The following local authorities have the smallest percentage of abstainers:12 Islesof Scilly, 6.3%; Mid Devon, 6.4%; and East Hampshire, 6.9%. Conversely, Newham, Brentand Tower Hamlets have the largest percentage of abstainers; 48.0%, 35.4% and 33.6%respectively. The local authorities with the smallest percentage of lower risk drinkers areNewham, 37.5%; Tower Hamlets, 43.6%; and Redbridge, 43.7%, while Tendring, 74.3%;Waveney, 74.1%; and Basildon, 73.5% have the largest percentage of lower risk drinkers.Considering increasing risk drinkers, Brent, 8.6%; Slough, 8.8%; and Ipswich, 9.1% have thesmallest proportion of these, while Exeter, 30.9%; Isles of Scilly, 30.8%; Mid Devon, 28.7%;and Teignbridge, 28.7% have the largest. Finally, the following local authorities have thesmallest percentage of higher risk drinkers: Peterborough, 2.0%; Weymouth and Portland,2.2%; and Kings Lynn and West Norfolk, 2.2%, while Harrow, 11.7%; Hounslow, 11.7%; andDoncaster, 11.2% have the highest. Maps 1 to 4 illustrate differences at the local authoritylevel in the estimated percentage of the adult population who are abstainers, lower risk,increasing risk and higher risk drinkers respectively.

The number of alcohol abstainers within a population affects the proportion of drinkers whoexceed government guidelines. If, for example, a number of lower risk drinkers within an areabecome abstainers, the proportion of all drinkers who exceed guidelines will rise, even thoughthe actual numbers who consume alcohol at increasing or higher risk levels remainsunchanged. Using Redbridge as an example, it is estimated that 16.4% of people inRedbridge are increasing risk drinkers, but when limited to the drinking population only, it isestimated that 24.5% of drinkers in Redbridge are increasing risk drinkers. The correspondingfigures for higher risk drinkers in Redbridge are 6.7% and 10.0% respectively. Largedifferences across local authorities are again observed. For example, 57.2% of adult drinkersin Exeter drink at lower risk levels, compared to 86.4% in Kings Lynn and West Norfolk. InIpswich 10.8% of adult drinkers are increasing risk drinkers, while in Exeter, this proportion is33.5%. An estimated 2.5% of adult drinkers in Peterborough are higher risk drinkerscompared to 16.1% in Hounslow. Maps 5 to 7, and Tables 2 to 10, detail the proportion ofadult drinkers who drink at lower risk, increasing risk and higher risk levels.

There are a number of reasons why these local authority estimates vary by area and acrosstime, but firstly, it is important to reflect upon the definitions for increasing risk and higher riskdrinking. Women are considered to be drinking at increasing risk levels if they regularly drinkmore than 14 units, up to 35 units, per week and are considered to be drinking at higher risk

1295% confidence intervals for all estimates are presented in Appendix 4.


7/68

6

levels if they regularly consume more than 35 units per week. If there are ten units of alcoholin a bottle of wine,13 for example, a woman who drinks half of a bottle of wine on three nightsof the week would be defined as an increasing risk drinker, while consuming half a bottle ofwine on four nights during the week and an additional two bottles during the course of theweekend, would push them into the higher risk category. Men are considered to be drinking atincreasing risk levels if they drink more than 21 units, up to 50 units, per week and are

considered to be drinking at higher risk levels if they regularly drink more than 50 units perweek. Therefore a man who drinks half of a bottle of wine on five nights of the week isdrinking at increasing risk levels, while a man who drinks two thirds of a bottle of wine oneach day of the week and an additional two bottles over the course of a weekend would beconsidered a higher risk drinker.

Changes over time are also, in part, due to a change in the survey (General Lifestyle Survey)methodology. The number of units of alcohol a person is reported to consume is based uponthe size of a given measure (for example, the size of a glass of wine) and the alcohol contentof the particular drink (the percentage of alcohol by volume; ABV). In recent years, both ofthese factors have changed and are reflected in the methodology employed by the GeneralLifestyle Survey from 2006 onwards. In relation to the ABV, the revised method changed thenumber of units assumed to be in a normal strength beer, lager and cider, strong beer, lagerand cider and wine categories. The size of a wine glass was also assumed to have changed.Until 2006, a wine glass was assumed to be 125ml; while in 2006 and 2007 it was assumedto be an average size of 170ml. The biggest impact of the 2006 methodological revision wastherefore on wine drinking, with a change in both the glass size and ABV of wine (from 9 to12%) resulting in the number of units assumed to be in a glass of wine doubling from one totwo units. In 2008, a new question about wine was included in the survey, with respondentsbeing asked whether they consumed a small (125 ml), standard (175 ml) or large (250 ml)glass of wine. Responses from this question were used when calculating the number of unitsconsumed; a small glass was assumed to contain one and a half units, a standard glass wasassumed to contain two units and a large glass was assumed to contain three units.14 Theimpact of the changes detailed above on estimates for alcohol consumption will be furtheraccentuated by changes in actual drinking behaviour, with figures from the British Beer andPub Association showing that, against a backdrop of falling beer consumption, the amount ofwine consumed per person by UK residents aged 15 and over has increased from 15.4 litresin 1990 to 25.6 litres in 2009.15

The previous estimates for increasing and higher risk drinking calculated by the North WestPublic Health Observatory used aggregated data from the General Lifestyle Survey for years2000 to 2002, while the current estimates used survey data from 2008. According to theGeneral Lifestyle Survey, the percentage of men drinking more than 21 units per week andthe percentage of women drinking more than 14 units per week fell between 2000 and 2006. 16However, data from the General Lifestyle Survey also show that between 2000/02 and 2008the proportion of men drinking at increasing risk levels has not changed considerably; 28% in2000/02 to 27% in 2008. The proportion of men drinking at higher risk levels has also notchanged; 7% in both 2000/02 and 2008. In 2000/02, just over 16% of women were increasing

13Improving accuracy in recording alcohol consumption: A survey in Greater Manchester. Morleo et al., 2011.

14General Lifestyle Survey, 2008. Smoking and Drinking Among Adults, 2008.Office for National Statistics, 2010.

15Statistical Handbook. A compilation of drinks industry statistics. British Beer and Pub Association, 2010.



8/68

7

risk drinkers compared to 19% in 2008, while the proportions who were higher risk drinkerswere 3% and 4% respectively.17 Clearly, the methodological changes detailed above makeinterpretation of trend data difficult. Furthermore, the impact of these methodological changesaffects areas differently depending upon the populations characteristics. Until 2006, theGeneral Lifestyle Survey showed a small difference in usual weekly alcohol consumptionbetween those in manual and non-manual households, with those in non-manual households

tending to have a higher weekly consumption. Following the methodological changes, agreater difference between manual and non-manual average weekly alcohol consumptionwas observed, a difference at least in part related to the relatively low prevalence of wineconsumption in routine and manual households compared with other categories.18 While theproportion of people who regularly exceed drinking guidelines are the most affluent, it isimportant to note that consumption per head is highest among men in lower paid employment.Such patterns of consumption may help explain why the poorest communities continue toexperience the highest rates of ill health, hospital admissions and death.19,20

Previous estimates of increasing and higher risk drinking produced by the North West PublicHealth Observatory were based upon aggregated General Lifestyle Survey data for years2000 to 2002. The new estimates presented here are based upon survey data for 2008 whichwere the most recent data available when the estimates were produced; data were notaggregated across years because the questions on average weekly alcohol consumptionwere included in the General Lifestyle Survey in 2006 and again in 2008, but not in 2007.21The changes to the methodology of the General Lifestyle Survey detailed above, coupled withchanges in wine consumption, particularly among those in more affluent populations, suggestthat the previous 2000/02 estimates may have underestimated alcohol consumptionespecially in some communities.



19Safe, Sensible, Social Consultation on further action. Department of Health 2008.

20Alcohol pen portraits: Segmentation series report 4. Morleo et al. 2010.



9/68

8

Map 1: Estimated percentage of people aged 16 and over who abstain from alcohol,by local authority


10/68

9

Map 2: Estimated percentage of people aged 16 and over who are lower risk drinkers,by local authority


11/68

10

Map 3: Estimated percentage of people aged 16 and over who are increasing risk drinkers,by local authority


12/68

11

Map 4: Estimated percentage of people aged 16 and over who are higher risk drinkers,by local authority


13/68

12

Map 5: Estimated percentage of drinkers aged 16 and over who are drinking at lower risklevels, by local authority


14/68

13

Map 6: Estimated percentage of drinkers aged 16 and over who are drinking at increasing risklevels, by local authority


15/68

14

Map 7: Estimated percentage of drinkers aged 16 and over who are drinking at higher risklevels, by local authority


16/68

15

Table 2: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the East Midlands, by local authority

East Midlands Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)

Local AuthorityAbstain Lower Increasing Higher Lower Increasing Higher Abstain Lower Increasing Higher

East Midlands 13.7% 62.3% 17.9% 6.1% 72.2% 20.7% 7.0% 492033 2231350 640180 217374

Amber Valley 11.1% 62.8% 20.1% 6.0% 70.6% 22.6% 6.7% 10881 61826 19815 5903

Ashfield 11.9% 64.9% 14.8% 8.4% 73.7% 16.8% 9.5% 11161 61006 13952 7852

Bassetlaw 13.8% 67.3% 14.2% 4.7% 78.1% 16.4% 5.5% 12625 61313 12913 4313

Blaby 11.6% 61.0% 20.4% 6.9% 69.1% 23.1% 7.8% 8773 46062 15387 5234

Bolsover 13.3% 63.3% 15.3% 8.1% 73.0% 17.6% 9.4% 8022 38179 9219 4897

Boston 12.9% 65.6% 14.7% 6.8% 75.3% 16.9% 7.8% 6163 31368 7026 3242

Broxtowe 13.2% 63.6% 16.8% 6.5% 73.2% 19.4% 7.4% 12209 58938 15584 5988

Charnwood 15.0% 61.6% 19.2% 4.3% 72.4% 22.5% 5.0% 20428 83934 26131 5846

Chesterfield 12.4% 66.2% 16.0% 5.3% 75.6% 18.3% 6.1% 10254 54592 13221 4378

Corby 14.1% 61.3% 19.6% 5.0% 71.3% 22.8% 5.9% 6173 26920 8608 2218

Daventry 8.6% 63.7% 21.9% 5.8% 69.7% 24.0% 6.4% 5444 40322 13878 3680

Derby 16.7% 61.3% 14.9% 7.0% 73.7% 17.9% 8.5% 31958 117340 28504 13475

Derbyshire Dales 11.2% 61.7% 19.7% 7.4% 69.5% 22.2% 8.3% 6491 35852 11424 4296

East Lindsey 13.5% 67.8% 12.7% 6.0% 78.4% 14.7% 6.9% 15885 79943 14993 7069

East Northamptonshire 10.3% 64.0% 20.8% 4.9% 71.4% 23.2% 5.4% 7003 43442 14120 3304

Erewash 11.3% 62.7% 20.0% 6.0% 70.7% 22.6% 6.7% 10192 56623 18067 5386

Gedling 11.0% 64.1% 18.2% 6.8% 71.9% 20.4% 7.6% 10086 58963 16732 6265

Harborough 10.8% 61.9% 22.3% 4.9% 69.5% 25.0% 5.5% 7156 40962 14749 3268

High Peak 11.1% 62.7% 20.2% 6.0% 70.5% 22.7% 6.8% 8369 47328 15250 4566

Hinckley and Bosworth 10.7% 63.8% 20.7% 4.8% 71.5% 23.1% 5.4% 9181 54798 17753 4138

Kettering 10.7% 63.9% 20.6% 4.8% 71.5% 23.1% 5.4% 7676 45855 14820 3454

Leicester 29.5% 52.4% 14.3% 3.7% 74.4% 20.3% 5.3% 68314 121477 33232 8666

Lincoln 13.9% 65.4% 15.1% 5.7% 75.9% 17.5% 6.6% 10046 47279 10918 4095

Mansfield 13.6% 62.1% 16.8% 7.5% 71.9% 19.4% 8.7% 11092 50567 13667 6096

Melton 9.9% 63.5% 21.2% 5.3% 70.6% 23.6% 5.9% 3987 25485 8518 2116

Newark and Sherwood 11.0% 62.2% 19.2% 7.5% 70.0% 21.6% 8.5% 10113 56976 17558 6893

North East Derbyshire 12.6% 64.2% 18.1% 5.1% 73.4% 20.7% 5.9% 10259 52165 14718 4160

North Kesteven 9.8% 63.4% 19.4% 7.4% 70.3% 21.5% 8.2% 8428 54563 16684 6391

North West Leicestershire 9.8% 60.3% 23.7% 6.1% 66.9% 26.3% 6.8% 7195 44148 17346 4469


17/68

16

East Midlands Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


East Midlands 13.7% 62.3% 17.9% 6.1% 72.2% 20.7% 7.0% 492033 2231350 640180 217374

Northampton 15.2% 65.0% 16.1% 3.7% 76.6% 19.0% 4.3% 24878 106275 26390 6021

Nottingham 18.5% 59.3% 15.3% 6.9% 72.8% 18.8% 8.5% 44213 141489 36482 16430

Oadby and Wigston 17.3% 56.7% 19.2% 6.9% 68.5% 23.2% 8.3% 7997 26234 8864 3188

Rushcliffe 9.0% 64.3% 18.3% 8.4% 70.6% 20.2% 9.2% 7997 57031 16271 7431

Rutland 12.7% 59.8% 21.3% 6.2% 68.5% 24.4% 7.0% 3934 18578 6630 1911

South Derbyshire 11.1% 62.4% 18.6% 7.9% 70.2% 20.9% 8.9% 8071 45513 13567 5781

South Holland 11.8% 61.3% 20.0% 6.9% 69.4% 22.7% 7.8% 8045 41841 13684 4722

South Kesteven 10.0% 64.9% 17.9% 7.3% 72.1% 19.8% 8.1% 10599 69036 18982 7712

South Northamptonshire 10.0% 62.2% 22.8% 5.1% 69.1% 25.3% 5.6% 7226 44844 16428 3648

Wellingborough 12.4% 56.2% 23.4% 7.9% 64.2% 26.7% 9.1% 7578 34241 14233 4829

West Lindsey 8.2% 66.8% 19.3% 5.6% 72.8% 21.0% 6.1% 5929 48045 13864 4043

Local authority values do not always sum to regional total due to rounding.


18/68

17

Table 3: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the East of England, by local authority

East of England Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


East of England 13.1% 68.4% 14.6% 4.0% 78.7% 16.8% 4.6% 600722 3131670 666916 181449

Babergh 11.0% 72.4% 14.2% 2.4% 81.4% 15.9% 2.7% 7749 50776 9954 1686

Basildon 10.6% 73.5% 11.0% 4.9% 82.2% 12.3% 5.5% 14344 99329 14820 6643

Bedford 18.5% 61.6% 17.3% 2.6% 75.6% 21.2% 3.2% 22933 76486 21438 3250

Braintree 8.8% 72.1% 15.1% 4.0% 79.1% 16.5% 4.4% 9970 81289 16989 4538

Breckland 13.4% 69.5% 14.3% 2.9% 80.2% 16.5% 3.3% 14285 74213 15273 3083

Brentwood 10.6% 68.4% 15.6% 5.4% 76.5% 17.5% 6.0% 6161 39667 9062 3119

Broadland 11.0% 70.8% 14.6% 3.6% 79.6% 16.4% 4.0% 11164 72046 14855 3657

Broxbourne 12.9% 64.6% 18.8% 3.8% 74.1% 21.5% 4.3% 9216 46264 13429 2705

Cambridge 16.8% 68.6% 12.3% 2.4% 82.4% 14.7% 2.9% 17239 70452 12591 2492

Castle Point 11.5% 68.6% 15.2% 4.7% 77.5% 17.2% 5.3% 8416 50136 11143 3425

Central Bedfordshire 14.0% 61.3% 19.5% 5.2% 71.2% 22.7% 6.1% 28261 123928 39469 10554

Chelmsford 9.7% 68.6% 16.2% 5.6% 75.9% 17.9% 6.2% 12874 91514 21542 7441

Colchester 9.6% 71.7% 14.8% 4.0% 79.3% 16.4% 4.4% 13787 103080 21270 5696

Dacorum 10.7% 65.4% 19.1% 4.7% 73.3% 21.4% 5.3% 11927 72666 21262 5212

East Cambridgeshire 13.3% 69.9% 13.8% 3.0% 80.6% 15.9% 3.5% 8684 45697 9022 1978

East Hertfordshire 9.9% 65.8% 19.5% 4.8% 73.1% 21.6% 5.3% 10609 70299 20817 5096

Epping Forest 12.9% 67.7% 14.8% 4.5% 77.8% 17.0% 5.2% 12886 67624 14781 4536

Fenland 14.2% 70.1% 13.3% 2.4% 81.7% 15.5% 2.8% 10605 52356 9909 1793

Forest Heath 12.3% 71.3% 13.1% 3.4% 81.2% 14.9% 3.9% 6192 36021 6597 1724

Great Yarmouth 15.7% 69.4% 11.1% 3.8% 82.3% 13.2% 4.5% 12097 53598 8564 2930

Harlow 13.8% 72.2% 10.5% 3.5% 83.8% 12.1% 4.1% 8634 45011 6525 2193

Hertsmere 14.4% 64.4% 17.4% 3.8% 75.2% 20.4% 4.4% 11121 49838 13492 2929

Huntingdonshi re 10.4% 68.5% 14.8% 6.3% 76.5% 16.5% 7.0% 14017 92299 19946 8438

Ipswich 16.3% 71.9% 9.1% 2.7% 85.9% 10.8% 3.3% 15893 70306 8879 2687

King's Lynn and West Norfolk 15.4% 73.1% 9.3% 2.2% 86.4% 11.0% 2.6% 18191 86520 10978 2636

Luton 25.0% 60.4% 12.0% 2.6% 80.6% 16.0% 3.4% 36722 88729 17566 3797

Maldon 9.0% 71.0% 16.1% 3.8% 78.1% 17.7% 4.2% 4542 35874 8144 1944

Mid Suffolk 10.7% 70.8% 14.8% 3.7% 79.3% 16.6% 4.1% 8175 54012 11294 2790

North Hertfordshire 13.1% 63.3% 19.6% 4.0% 72.8% 22.5% 4.6% 12848 62241 19250 3947


19/68

18

East of England Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


East of England 13.1% 68.4% 14.6% 4.0% 78.7% 16.8% 4.6% 600722 3131670 666916 181449

North Norfolk 13.8% 70.4% 12.9% 2.9% 81.7% 15.0% 3.4% 11835 60138 11018 2493

Norwich 16.0% 70.6% 10.9% 2.6% 84.0% 12.9% 3.1% 17863 78821 12122 2900

Peterborough 18.1% 69.9% 10.0% 2.0% 85.4% 12.2% 2.5% 23375 90291 12868 2610

Rochford 11.2% 66.4% 15.5% 6.9% 74.8% 17.4% 7.8% 7464 44229 10320 4603

South Cambridgeshire 12.9% 66.7% 15.1% 5.3% 76.5% 17.3% 6.1% 14168 73225 16579 5870

South Norfolk 11.1% 70.8% 14.5% 3.6% 79.6% 16.4% 4.0% 10673 68099 13988 3454

Southend-on-Sea 12.8% 69.5% 11.7% 5.9% 79.8% 13.5% 6.8% 16786 91117 15385 7728

St Albans 11.1% 65.3% 18.9% 4.6% 73.5% 21.3% 5.2% 11603 67948 19723 4816

St Edmundsbury 12.3% 69.1% 15.0% 3.6% 78.8% 17.1% 4.1% 10324 57809 12539 3036

Stevenage 12.5% 65.8% 18.9% 2.8% 75.2% 21.6% 3.2% 7862 41486 11947 1767

Suffolk Coastal 11.5% 72.1% 14.0% 2.4% 81.5% 15.8% 2.7% 11753 73602 14250 2411

Tendring 10.1% 74.3% 11.7% 4.0% 82.6% 13.0% 4.4% 12278 90754 14239 4874

Three Rivers 14.3% 62.3% 17.8% 5.6% 72.6% 20.8% 6.6% 9875 43020 12308 3892

Thurrock 13.0% 72.0% 10.5% 4.5% 82.8% 12.1% 5.1% 15358 85047 12395 5268

Uttlesford 11.2% 66.1% 15.7% 7.0% 74.4% 17.6% 7.9% 6428 38112 9030 4057

Watford 15.6% 59.5% 20.4% 4.5% 70.5% 24.1% 5.4% 9981 38051 13012 2896

Waveney 11.8% 74.1% 10.4% 3.7% 84.0% 11.8% 4.2% 11413 71543 10091 3556

Welwyn Hatfield 14.0% 64.6% 18.7% 2.7% 75.2% 21.8% 3.1% 12139 56109 16241 2301



20/68

19

Table 4: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in London, by local authority

London Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


London 24.5% 52.1% 15.8% 7.6% 69.1% 20.9% 10.0% 1496900 3178006 960707 462339

Barking and Dagenham 28.1% 49.0% 16.1% 6.9% 68.1% 22.4% 9.5% 35539 62031 20362 8674

Barnet 28.6% 53.7% 13.6% 4.1% 75.2% 19.1% 5.7% 75229 141054 35771 10654

Bexley 11.2% 58.4% 23.8% 6.6% 65.8% 26.8% 7.4% 19826 103720 42317 11662

Brent 35.4% 48.7% 8.6% 7.3% 75.4% 13.3% 11.4% 77507 106721 18776 16095

Bromley 14.1% 53.2% 24.1% 8.5% 62.0% 28.1% 9.9% 33975 128601 58330 20650

Camden 24.5% 50.4% 16.5% 8.6% 66.8% 21.8% 11.4% 47973 98684 32238 16794

City of London 19.3% 50.4% 22.4% 7.9% 62.5% 27.7% 9.8% 1400 3661 1622 574

Croydon 28.5% 53.7% 12.8% 5.0% 75.1% 17.8% 7.0% 76701 144562 34358 13561

Ealing 27.1% 55.1% 10.2% 7.6% 75.6% 13.9% 10.5% 67078 136205 25141 18913

Enfield 22.7% 56.8% 14.7% 5.8% 73.4% 19.0% 7.5% 51083 127622 33095 13067

Greenwich 20.3% 55.1% 16.9% 7.6% 69.2% 21.3% 9.5% 35842 97171 29871 13372

Hackney 33.0% 45.7% 14.1% 7.2% 68.2% 21.1% 10.7% 53979 74618 23054 11680

Hammersmith and Fulham 20.6% 56.5% 13.5% 9.3% 71.2% 17.0% 11.8% 29738 81686 19532 13502

Haringey 30.7% 52.8% 13.2% 3.3% 76.2% 19.0% 4.8% 55522 95522 23842 6002

Harrow 26.2% 49.8% 12.3% 11.7% 67.5% 16.7% 15.9% 45161 85923 21250 20212

Havering 19.6% 46.4% 23.1% 10.9% 57.7% 28.8% 13.5% 36291 85836 42794 20104

Hillingdon 21.1% 53.3% 15.2% 10.3% 67.6% 19.3% 13.0% 42133 106334 30355 20497

Hounslow 27.7% 49.8% 10.9% 11.7% 68.9% 15.0% 16.1% 49031 88322 19288 20658

Islington 24.9% 51.7% 17.8% 5.6% 68.9% 23.7% 7.4% 39130 81369 28005 8769

Kensington and Chelsea 18.2% 61.6% 11.7% 8.4% 75.4% 14.3% 10.3% 27455 92949 17654 12721

Kingston upon Thames 15.0% 57.3% 20.9% 6.8% 67.4% 24.6% 8.0% 19417 73906 26935 8795

Lambeth 21.9% 55.2% 17.0% 5.9% 70.7% 21.7% 7.6% 49236 123855 38064 13358

Lewisham 22.2% 57.0% 15.4% 5.5% 73.2% 19.7% 7.0% 46029 118194 31873 11365

Merton 22.5% 51.4% 19.3% 6.8% 66.3% 24.9% 8.8% 36672 83581 31418 11044

Newham 48.0% 37.5% 10.2% 4.3% 72.1% 19.7% 8.2% 91407 71338 19460 8104

Redbridge 33.2% 43.7% 16.4% 6.7% 65.4% 24.5% 10.0% 66343 87191 32669 13377

Richmond upon Thames 12.1% 55.5% 22.0% 10.4% 63.1% 25.1% 11.8% 17528 80260 31900 15019

Southwark 20.5% 53.5% 17.3% 8.6% 67.4% 21.8% 10.8% 46229 120442 39023 19270

Sutton 17.2% 51.7% 22.4% 8.7% 62.5% 27.0% 10.5% 25571 76965 33275 12985


21/68

20

London Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


London 24.5% 52.1% 15.8% 7.6% 69.1% 20.9% 10.0% 1496900 3178006 960707 462339

Tower Hamlets 33.6% 43.6% 13.4% 9.4% 65.7% 20.1% 14.1% 57562 74594 22848 16059

Waltham Forest 32.7% 47.2% 13.3% 6.8% 70.2% 19.8% 10.1% 57306 82610 23257 11885

Wandsworth 15.8% 53.0% 22.0% 9.2% 62.9% 26.1% 10.9% 37493 125883 52301 21873

Westminster 22.4% 57.4% 9.9% 10.4% 73.9% 12.7% 13.3% 45513 116595 20031 21046



22/68

21

Table 5: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the North East, by local authority

North East Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


North East 16.6% 56.4% 20.5% 6.5% 67.6% 24.6% 7.8% 348405 1182784 430327 137258

County Durham 13.5% 58.1% 24.0% 4.4% 67.2% 27.7% 5.1% 56209 241191 99553 18340

Darlington 12.6% 59.8% 23.1% 4.5% 68.4% 26.4% 5.1% 10169 48262 18653 3621

Gateshead 19.3% 54.4% 18.1% 8.2% 67.4% 22.4% 10.2% 30144 85088 28287 12831

Hartlepool 13.5% 56.0% 24.2% 6.3% 64.7% 28.0% 7.3% 9832 40857 17696 4621

Middlesbrough 15.7% 54.8% 23.3% 6.1% 65.1% 27.7% 7.3% 17395 60702 25816 6768

Newcastle upon Tyne 21.7% 52.9% 17.4% 8.0% 67.6% 22.3% 10.2% 48909 119165 39230 17908

North Tyneside 17.4% 56.4% 17.7% 8.5% 68.3% 21.5% 10.3% 28038 90890 28560 13655

Northumberland 17.0% 59.0% 17.9% 6.1% 71.2% 21.5% 7.3% 43746 151526 45817 15547

Redcar and Cleveland 14.0% 58.0% 23.6% 4.4% 67.4% 27.5% 5.1% 15914 65651 26790 4931

South Tyneside 20.4% 53.8% 17.8% 8.1% 67.5% 22.3% 10.1% 25239 66646 22031 10000

Stockton-on-Tees 12.6% 57.4% 23.5% 6.5% 65.7% 26.9% 7.4% 19298 87764 35881 9939

Sunderland 18.9% 54.4% 18.3% 8.3% 67.2% 22.6% 10.3% 43511 125042 42012 19095



23/68

22

Table 6: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the North West, by local authority

North West Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


North West 14.7% 59.7% 19.3% 6.3% 70.0% 22.7% 7.3% 818514 3314340 1073429 347807

Allerdale 11.6% 62.9% 18.8% 6.8% 71.1% 21.2% 7.6% 8999 48981 14608 5255

Barrow-in-Furness 12.7% 57.3% 21.3% 8.7% 65.7% 24.4% 9.9% 7422 33380 12407 5039

Blackburn with Darwen 20.9% 53.0% 18.5% 7.6% 66.9% 23.4% 9.6% 22312 56509 19776 8123

Blackpool 13.6% 57.1% 20.9% 8.4% 66.1% 24.2% 9.8% 15879 66405 24277 9827

Bolton 16.7% 58.4% 19.1% 5.8% 70.1% 23.0% 7.0% 34706 121257 39706 12033

Burnley 16.1% 55.6% 20.1% 8.2% 66.2% 24.0% 9.8% 11193 38681 14020 5714

Bury 13.5% 63.3% 19.0% 4.3% 73.1% 21.9% 4.9% 19726 92571 27741 6260

Carlisle 11.6% 62.9% 18.8% 6.8% 71.1% 21.2% 7.6% 9909 53823 16083 5787

Cheshire East 14.7% 64.0% 16.9% 4.4% 75.0% 19.8% 5.1% 43092 188260 49799 12909

Cheshire West and Chester 13.1% 60.1% 19.7% 7.1% 69.2% 22.7% 8.1% 35144 160816 52690 18939

Chorley 10.4% 55.6% 27.7% 6.3% 62.0% 30.9% 7.1% 8863 47181 23499 5376

Copeland 12.8% 62.9% 18.9% 5.4% 72.1% 21.7% 6.2% 7387 36378 10940 3140

Eden 9.1% 60.6% 24.5% 5.9% 66.6% 26.9% 6.5% 3913 26063 10542 2525

Fylde 9.7% 55.7% 24.7% 10.0% 61.6% 27.3% 11.0% 6228 35840 15905 6420

Halton 15.2% 61.9% 16.7% 6.3% 72.9% 19.6% 7.4% 14453 58951 15877 6021

Hyndburn 15.8% 55.6% 20.3% 8.3% 66.1% 24.1% 9.8% 10182 35781 13049 5309

Knowsley 15.7% 61.7% 16.4% 6.2% 73.2% 19.4% 7.4% 18810 74077 19677 7474

Lancaster 12.6% 61.1% 20.2% 6.1% 69.9% 23.1% 7.0% 14961 72787 24058 7300

Liverpool 17.5% 60.6% 15.8% 6.0% 73.5% 19.2% 7.3% 62814 217288 56667 21536

Manchester 19.5% 57.1% 17.9% 5.4% 71.0% 22.3% 6.8% 72983 213388 66888 20298

Oldham 17.9% 57.7% 18.7% 5.7% 70.2% 22.8% 6.9% 30612 98530 32015 9730

Pendle 18.3% 57.9% 17.7% 6.1% 70.9% 21.6% 7.4% 13086 41405 12628 4333

Preston 17.7% 54.7% 19.6% 8.0% 66.4% 23.8% 9.7% 18805 58104 20860 8520

Ribble Valley 11.4% 55.9% 21.6% 11.1% 63.1% 24.4% 12.5% 5420 26483 10231 5240

Rochdale 17.1% 58.0% 19.1% 5.8% 70.0% 23.0% 7.0% 27849 94218 30943 9389

Rossendale 14.0% 62.1% 18.5% 5.3% 72.2% 21.6% 6.2% 7477 33097 9884 2846

Salford 14.3% 59.8% 19.9% 6.0% 69.8% 23.2% 7.0% 25547 106751 35535 10730

Sefton 14.8% 63.4% 15.6% 6.2% 74.4% 18.3% 7.3% 33523 143400 35235 14095

South Lakeland 9.0% 58.1% 23.5% 9.5% 63.8% 25.8% 10.4% 7863 50955 20596 8345


24/68

23

North West Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


North West 14.7% 59.7% 19.3% 6.3% 70.0% 22.7% 7.3% 818514 3314340 1073429 347807

South Ribble 10.6% 56.4% 26.5% 6.5% 63.1% 29.6% 7.3% 9197 49116 23068 5683

St Helens 15.6% 61.8% 16.4% 6.2% 73.2% 19.5% 7.4% 22379 88674 23589 8939

Stockport 11.7% 57.8% 23.7% 6.8% 65.4% 26.9% 7.7% 26581 131897 54158 15472

Tameside 14.6% 61.5% 19.7% 4.2% 72.1% 23.0% 4.9% 25179 106098 33893 7230

Trafford 13.1% 59.4% 22.9% 4.6% 68.4% 26.3% 5.3% 22487 101610 39154 7886

Warrington 13.0% 60.9% 21.6% 4.6% 69.9% 24.8% 5.3% 20396 95727 33934 7240

West Lancashire 11.5% 60.1% 18.8% 9.6% 67.9% 21.2% 10.9% 10226 53416 16720 8583

Wigan 12.6% 60.5% 20.7% 6.3% 69.2% 23.6% 7.1% 31066 149790 51098 15466

Wirral 16.5% 61.5% 16.0% 6.0% 73.6% 19.2% 7.2% 41282 154119 40145 15158

Wyre 11.4% 56.9% 23.3% 8.3% 64.3% 26.4% 9.3% 10563 52529 21536 7638



25/68

24

Table 7: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the South East, by local authority

South East Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


South East 12.1% 63.7% 18.3% 5.9% 72.5% 20.8% 6.7% 810924 4284160 1231994 396507

Adur 12.9% 66.7% 16.1% 4.3% 76.6% 18.5% 5.0% 6417 33159 8001 2148

Arun 10.8% 60.6% 23.7% 5.0% 67.9% 26.5% 5.6% 13181 74231 28975 6104

Ashford 11.3% 70.6% 14.5% 3.7% 79.6% 16.3% 4.1% 10076 62760 12870 3248

Aylesbury Vale 12.4% 63.6% 17.1% 6.9% 72.6% 19.6% 7.8% 17207 88152 23759 9524

Basingstoke and Deane 7.2% 66.8% 20.4% 5.6% 72.0% 21.9% 6.0% 9239 85494 26046 7137

Bracknell Forest 14.6% 60.5% 16.4% 8.5% 70.8% 19.2% 10.0% 13136 54537 14817 7664

Brighton and Hove 11.9% 61.4% 20.0% 6.7% 69.7% 22.7% 7.6% 25261 130372 42358 14277

Canterbury 12.0% 67.7% 14.5% 5.8% 76.9% 16.5% 6.6% 14652 82969 17808 7075

Cherwell 13.8% 64.3% 17.8% 4.1% 74.6% 20.7% 4.7% 15223 70674 19575 4496

Chichester 8.5% 61.2% 23.0% 7.3% 66.9% 25.1% 8.0% 7674 55376 20773 6618

Chiltern 13.0% 63.6% 16.7% 6.7% 73.1% 19.2% 7.7% 9367 45725 12040 4803

Crawley 11.5% 62.1% 22.0% 4.3% 70.2% 24.9% 4.9% 9235 49749 17622 3477

Dartford 12.8% 69.1% 14.6% 3.4% 79.3% 16.8% 3.9% 9307 50080 10614 2478

Dover 12.7% 67.4% 14.4% 5.4% 77.3% 16.5% 6.2% 11047 58482 12498 4684

East Hampshire 6.9% 67.2% 20.3% 5.6% 72.2% 21.8% 6.0% 6171 59779 18054 4955

Eastbourne 12.3% 63.2% 17.6% 7.0% 72.0% 20.0% 8.0% 9730 49989 13888 5518

Eastleigh 8.6% 65.5% 21.1% 4.8% 71.7% 23.0% 5.3% 8346 63667 20476 4697

Elmbridge 9.9% 61.5% 21.9% 6.8% 68.2% 24.3% 7.5% 10297 64186 22844 7090

Epsom and Ewell 11.4% 56.8% 22.3% 9.5% 64.1% 25.2% 10.7% 6492 32513 12766 5428

Fareham 10.2% 63.6% 20.2% 6.0% 70.8% 22.5% 6.7% 9143 57096 18168 5399

Gosport 9.5% 63.0% 22.0% 5.5% 69.6% 24.3% 6.1% 6077 40405 14089 3556

Gravesham 16.8% 66.1% 12.7% 4.4% 79.5% 15.3% 5.2% 13128 51646 9947 3402

Guildford 10.9% 58.0% 22.9% 8.2% 65.1% 25.7% 9.2% 12034 64118 25334 9063

Hart 10.2% 63.2% 20.5% 6.1% 70.4% 22.8% 6.8% 7273 45298 14686 4380

Hastings 12.9% 62.2% 20.3% 4.6% 71.4% 23.3% 5.3% 8970 43197 14101 3187

Havant 10.3% 71.9% 14.4% 3.4% 80.1% 16.1% 3.8% 9827 68782 13828 3269

Horsham 8.8% 60.7% 23.9% 6.6% 66.5% 26.2% 7.2% 9146 63188 24927 6881

Isle of Wight 9.7% 73.0% 14.2% 3.1% 80.9% 15.7% 3.4% 11240 84714 16479 3577

Lewes 8.7% 61.1% 22.9% 7.3% 67.0% 25.1% 8.0% 6778 47562 17796 5677


26/68

25



South East 12.1% 63.7% 18.3% 5.9% 72.5% 20.8% 6.7% 810924 4284160 1231994 396507

Maidstone 11.3% 69.9% 13.5% 5.3% 78.8% 15.3% 5.9% 13219 81589 15793 6134

Medway 13.6% 67.0% 15.0% 4.3% 77.6% 17.4% 5.0% 27313 134087 30120 8654

Mid Sussex 10.7% 58.3% 22.9% 8.2% 65.2% 25.6% 9.1% 11133 60628 23818 8497

Milton Keynes 15.2% 60.6% 17.8% 6.5% 71.4% 21.0% 7.6% 27244 108760 31910 11622

Mole Valley 7.7% 62.2% 22.6% 7.4% 67.4% 24.5% 8.1% 5080 40894 14890 4889

New Forest 7.4% 67.7% 19.7% 5.3% 73.1% 21.2% 5.7% 10726 98166 28516 7686

Oxford 16.9% 58.0% 18.4% 6.7% 69.8% 22.1% 8.1% 21462 73742 23397 8527

Portsmouth 11.9% 65.9% 17.3% 4.9% 74.8% 19.6% 5.6% 19462 107905 28332 8007

Reading 17.9% 59.2% 17.2% 5.6% 72.2% 20.9% 6.8% 21136 69773 20243 6610

Reigate and Banstead 9.5% 61.6% 22.1% 6.8% 68.0% 24.4% 7.6% 10097 65725 23608 7310

Rother 11.1% 59.1% 23.9% 5.9% 66.5% 26.9% 6.6% 8217 43564 17618 4332

Runnymede 11.4% 57.9% 22.6% 8.1% 65.4% 25.5% 9.1% 7885 39944 15592 5561

Rushmoor 11.0% 65.0% 20.7% 3.2% 73.1% 23.3% 3.6% 7870 46349 14776 2259

Sevenoaks 11.9% 68.3% 14.5% 5.3% 77.5% 16.4% 6.0% 10887 62751 13302 4872

Shepway 14.6% 70.3% 10.1% 5.0% 82.4% 11.8% 5.8% 11951 57559 8239 4088

Slough 33.4% 54.9% 8.8% 2.9% 82.4% 13.3% 4.4% 31358 51606 8312 2746

South Bucks 14.1% 63.0% 16.4% 6.5% 73.3% 19.1% 7.6% 7287 32469 8448 3363

South Oxfordshire 13.9% 62.2% 17.8% 6.0% 72.3% 20.7% 6.9% 14328 64023 18361 6151

Southampton 11.1% 67.2% 16.3% 5.4% 75.6% 18.4% 6.1% 21356 129662 31519 10413

Spelthorne 9.9% 58.1% 24.3% 7.8% 64.4% 26.9% 8.7% 7302 43050 17990 5788

Surrey Heath 10.1% 56.0% 24.3% 9.6% 62.3% 27.0% 10.7% 6701 37134 16093 6372

Swale 13.3% 70.9% 10.5% 5.3% 81.8% 12.1% 6.1% 13840 73624 10925 5467

Tandridge 7.9% 62.0% 22.6% 7.4% 67.4% 24.6% 8.1% 5247 40961 14931 4916

Test Valley 8.4% 65.6% 21.1% 4.9% 71.7% 23.0% 5.3% 7760 60492 19429 4471

Thanet 17.0% 66.0% 11.9% 5.1% 79.5% 14.4% 6.1% 17784 68937 12484 5278

Tonbridge and Malling 11.0% 70.1% 13.6% 5.3% 78.8% 15.3% 6.0% 10113 64378 12490 4872

Tunbridge Wells 10.8% 68.7% 14.4% 6.2% 77.0% 16.1% 6.9% 9033 57505 12045 5154

Vale of White Horse 14.0% 62.2% 17.8% 6.0% 72.3% 20.7% 6.9% 13066 58251 16703 5593

Waverley 9.1% 59.8% 21.8% 9.3% 65.8% 24.0% 10.2% 8487 56049 20457 8682


27/68

26



South East 12.1% 63.7% 18.3% 5.9% 72.5% 20.8% 6.7% 810924 4284160 1231994 396507

Wealden 10.8% 61.1% 22.5% 5.5% 68.5% 25.3% 6.2% 12677 71589 26399 6458

West Berkshire 13.3% 63.3% 16.9% 6.4% 73.1% 19.5% 7.4% 15943 75768 20271 7663

West Oxfordshire 13.6% 62.5% 18.0% 6.0% 72.3% 20.8% 6.9% 11089 51102 14715 4907

Winchester 7.3% 67.2% 20.0% 5.5% 72.5% 21.6% 5.9% 6535 60476 18032 4936

Windsor and Maidenhead 13.7% 63.1% 16.6% 6.6% 73.1% 19.2% 7.7% 15443 71088 18713 7449

Woking 9.7% 57.0% 22.7% 10.6% 63.2% 25.1% 11.7% 7143 41794 16608 7763

Wokingham 15.0% 61.4% 17.6% 5.9% 72.3% 20.7% 7.0% 18803 76782 22017 7430

Worthing 10.9% 60.3% 23.8% 5.0% 67.6% 26.7% 5.6% 8926 49344 19485 4113

Wycombe 16.7% 61.5% 15.8% 6.0% 73.8% 19.0% 7.2% 21350 78743 20274 7665



28/68

27

Table 8: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the South West, by local authority

South West Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


South West 11.8% 63.7% 19.8% 4.7% 72.2% 22.4% 5.4% 503271 2708610 841012 201012

Bath and North East Somerset 9.0% 64.8% 21.5% 4.7% 71.2% 23.6% 5.1% 13245 95829 31807 6892

Bournemouth 16.0% 65.2% 13.3% 5.5% 77.6% 15.8% 6.6% 22039 89702 18248 7621

Bristol, City of 14.3% 64.8% 17.2% 3.7% 75.6% 20.1% 4.3% 49447 223969 59562 12780

Cheltenham 10.7% 65.6% 20.7% 3.0% 73.5% 23.2% 3.3% 9910 60546 19134 2741

Christchurch 13.6% 66.1% 16.5% 3.9% 76.4% 19.1% 4.5% 5195 25314 6317 1483

Cornwall 11.4% 63.4% 20.2% 5.0% 71.5% 22.8% 5.7% 49918 277528 88555 22043

Cotswold 10.8% 64.8% 19.7% 4.7% 72.6% 22.1% 5.3% 7414 44661 13615 3251

East Devon 8.4% 60.5% 26.0% 5.1% 66.1% 28.4% 5.5% 9322 67314 28938 5620

East Dorset 14.7% 65.7% 16.3% 3.3% 77.0% 19.0% 3.9% 10522 47085 11642 2387

Exeter 7.8% 52.8% 30.9% 8.5% 57.2% 33.5% 9.2% 8094 54693 32038 8813

Forest of Dean 10.2% 67.3% 19.2% 3.3% 74.9% 21.4% 3.7% 6809 45083 12889 2218

Gloucester 12.0% 70.0% 15.1% 2.9% 79.5% 17.2% 3.3% 10976 63930 13818 2653

Isles of Scilly 6.3% 54.6% 30.8% 8.3% 58.3% 32.8% 8.9% 109 945 532 144

Mendip 15.4% 64.4% 16.7% 3.5% 76.1% 19.7% 4.2% 13505 56576 14677 3119

Mid Devon 6.4% 59.3% 28.7% 5.5% 63.4% 30.7% 5.9% 3940 36524 17700 3409

North Devon 8.7% 58.1% 27.7% 5.5% 63.6% 30.3% 6.0% 6559 43909 20937 4144

North Dorset 16.6% 64.7% 15.7% 2.9% 77.6% 18.9% 3.5% 9105 35386 8613 1591

North Somerset 11.3% 67.2% 18.9% 2.6% 75.8% 21.3% 2.9% 18932 112585 31683 4299

Plymouth 11.2% 63.1% 20.6% 5.2% 71.1% 23.1% 5.8% 23054 130275 42425 10636

Poole 15.9% 65.0% 16.3% 2.8% 77.3% 19.4% 3.3% 18069 73845 18499 3197

Purbeck 16.7% 65.7% 14.5% 3.0% 78.9% 17.5% 3.6% 6402 25125 5561 1154

Sedgemoor 13.7% 63.6% 18.5% 4.1% 73.7% 21.5% 4.8% 12566 58247 16989 3800

South Gloucestershire 9.7% 64.2% 20.1% 6.0% 71.1% 22.2% 6.7% 20112 132914 41576 12442

South Hams 9.6% 56.7% 27.5% 6.2% 62.7% 30.4% 6.9% 6637 39357 19117 4339

South Somerset 15.2% 65.1% 16.0% 3.7% 76.8% 18.9% 4.3% 19628 84004 20649 4714

Stroud 8.6% 68.6% 20.2% 2.6% 75.1% 22.1% 2.8% 7728 61475 18118 2308

Swindon 15.8% 60.6% 17.6% 6.0% 71.9% 20.9% 7.2% 24049 92046 26759 9154

Taunton Deane 14.2% 66.6% 16.0% 3.2% 77.6% 18.6% 3.8% 12505 58706 14070 2839

Teignbridge 7.6% 55.8% 28.7% 7.9% 60.4% 31.1% 8.5% 8015 58591 30160 8248


29/68

28

South West Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


South West 11.8% 63.7% 19.8% 4.7% 72.2% 22.4% 5.4% 503271 2708610 841012 201012

Tewkesbury 9.7% 65.4% 21.0% 3.8% 72.5% 23.3% 4.2% 6302 42489 13659 2479

Torbay 10.5% 59.4% 21.9% 8.1% 66.4% 24.5% 9.1% 11738 66256 24449 9060

Torridge 11.0% 61.8% 21.3% 5.9% 69.4% 23.9% 6.7% 5923 33240 11449 3185

West Devon 9.6% 55.9% 28.5% 6.0% 61.9% 31.5% 6.6% 4122 24017 12229 2567

West Dorset 14.8% 67.6% 15.3% 2.3% 79.3% 18.0% 2.7% 11898 54397 12343 1820

West Somerset 16.2% 69.9% 11.0% 2.9% 83.4% 13.1% 3.5% 4899 21171 3332 890

Weymouth and Portland 18.7% 68.1% 11.0% 2.2% 83.8% 13.5% 2.7% 10081 36774 5912 1200

Wiltshire 9.5% 64.4% 20.1% 6.0% 71.2% 22.2% 6.6% 34500 234103 73010 21770



30/68

29

Table 9: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in the West Midlands, by local authority

West Midlands Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)

Local Authority Abstain Lower Increasing Higher Lower Increasing Higher Abstain Lower Increasing Higher

West Midlands 17.3% 63.5% 14.9% 4.3% 76.7% 18.1% 5.2% 747399 2748094 646774 186636

Birmingham 27.6% 58.4% 11.2% 2.8% 80.7% 15.4% 3.9% 217152 459743 87868 22188

Bromsgrove 11.7% 61.8% 18.4% 8.1% 70.0% 20.8% 9.2% 8778 46511 13808 6119

Cannock Chase 10.5% 68.8% 14.6% 6.0% 76.9% 16.4% 6.7% 7999 52334 11130 4571

Coventry 22.7% 59.7% 12.9% 4.7% 77.2% 16.7% 6.1% 56028 147251 31892 11629

Dudley 16.1% 69.5% 11.8% 2.6% 82.9% 14.0% 3.1% 39865 171939 29113 6462

East Staffordshire 10.5% 65.2% 19.0% 5.2% 72.9% 21.3% 5.8% 9161 56734 16539 4544

Herefordshire, County of 11.8% 65.7% 17.9% 4.6% 74.5% 20.3% 5.2% 17388 96592 26357 6704

Lichfie ld 10.6% 67.7% 16.9% 4.7% 75.8% 18.9% 5.3% 8482 53961 13430 3778

Malvern Hills 11.5% 64.8% 18.2% 5.5% 73.3% 20.5% 6.2% 6994 39513 11067 3360

Newcastle-under-Lyme 10.3% 64.0% 19.7% 6.1% 71.3% 21.9% 6.8% 10548 65871 20242 6246

North Warwickshire 13.0% 63.1% 18.8% 5.0% 72.6% 21.6% 5.8% 6640 32145 9563 2562

Nuneaton and Bedworth 15.7% 64.9% 14.0% 5.4% 77.0% 16.6% 6.4% 15314 63118 13651 5238

Redditch 15.3% 65.9% 15.4% 3.5% 77.7% 18.2% 4.1% 9749 42008 9808 2216

Rugby 12.5% 63.9% 18.1% 5.5% 73.0% 20.7% 6.3% 9070 46542 13172 4036

Sandwell 24.2% 60.7% 12.1% 3.0% 80.1% 15.9% 4.0% 55061 138412 27528 6906

Shropshire 10.6% 67.4% 17.4% 4.5% 75.4% 19.5% 5.1% 25298 160932 41637 10781

Solihull 13.1% 66.8% 16.3% 3.8% 76.9% 18.8% 4.4% 21423 109559 26756 6243

South Staffordshire 9.5% 66.3% 18.6% 5.5% 73.3% 20.6% 6.1% 8396 58352 16355 4867

Stafford 8.8% 64.9% 19.9% 6.5% 71.1% 21.8% 7.1% 9025 66586 20396 6669

Staffordshire Moorlands 10.2% 67.5% 17.7% 4.6% 75.2% 19.7% 5.1% 8108 53420 13985 3620

Stoke-on-Trent 13.6% 65.2% 16.0% 5.2% 75.5% 18.5% 6.0% 26348 126230 30984 10065

Stratford-on-Avon 14.0% 62.8% 16.8% 6.4% 73.1% 19.5% 7.4% 13527 60601 16184 6149

Tamworth 8.9% 63.3% 20.0% 7.8% 69.5% 21.9% 8.5% 5387 38128 12024 4681

Telford and Wrekin 10.2% 68.6% 15.1% 6.1% 76.4% 16.8% 6.8% 13002 87830 19331 7807

Walsall 21.2% 62.7% 12.9% 3.2% 79.6% 16.3% 4.1% 42677 126194 25888 6487

Warwick 14.3% 62.9% 17.9% 4.9% 73.3% 20.9% 5.7% 15950 70062 19974 5488

Wolverhampton 24.6% 61.4% 10.8% 3.2% 81.4% 14.3% 4.3% 46695 116341 20464 6107

Worcester 13.4% 63.3% 19.4% 3.9% 73.1% 22.4% 4.5% 10132 47986 14688 2963

Wychavon 12.7% 64.0% 18.0% 5.2% 73.4% 20.7% 5.9% 12242 61529 17314 4987


31/68

30

West Midlands Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


West Midlands 17.3% 63.5% 14.9% 4.3% 76.7% 18.1% 5.2% 747399 2748094 646774 186636

Wyre Forest 13.5% 63.5% 19.2% 3.9% 73.3% 22.2% 4.5% 10961 51671 15627 3165



32/68

31

Table 10: Estimates of abstainers, lower risk, increasing risk and higher risk drinkers in Yorkshire and Humber,by local authority

Yorkshire and Humber Population estimate for all groups Population estimate for drinkers only Population estimate per drinking group (N)


Yorkshire and Humber 14.3% 58.6% 18.5% 8.6% 68.4% 21.6% 10.0% 601407 2463157 776496 359314

Barnsley 14.4% 61.2% 18.1% 6.4% 71.5% 21.1% 7.4% 26233 111212 32836 11555

Bradford 22.0% 49.7% 19.9% 8.3% 63.8% 25.5% 10.7% 84976 191647 76681 32116

Calderdale 14.3% 58.3% 20.0% 7.4% 68.0% 23.4% 8.6% 22936 93361 32071 11849

Craven 12.4% 61.9% 18.5% 7.1% 70.7% 21.2% 8.1% 5757 28655 8588 3300

Doncaster 13.3% 57.7% 17.8% 11.2% 66.6% 20.5% 12.9% 31330 135621 41695 26213

East Riding of Yorkshire 11.0% 64.3% 17.8% 6.8% 72.3% 20.0% 7.7% 30398 177045 48928 18843

Hambleton 11.9% 58.2% 19.7% 10.2% 66.0% 22.4% 11.6% 8472 41575 14089 7324

Harrogate 11.0% 61.3% 17.7% 9.9% 68.9% 19.9% 11.1% 14345 79947 23112 12932

Kingston upon Hull, City of 15.8% 63.0% 16.4% 4.9% 74.7% 19.5% 5.8% 32927 131493 34216 10204

Kirklees 14.5% 55.0% 20.1% 10.4% 64.3% 23.5% 12.1% 46283 175019 64023 33010

Leeds 14.5% 55.5% 19.7% 10.3% 64.9% 23.0% 12.1% 90595 347789 123345 64622

North East Lincolnshire 14.9% 64.0% 14.3% 6.7% 75.2% 16.9% 7.9% 18983 81345 18231 8536

North Lincolnshire 13.6% 65.6% 13.9% 6.9% 75.9% 16.1% 8.0% 17581 84764 17985 8902

Richmondshire 12.7% 62.3% 18.8% 6.1% 71.4% 21.6% 7.0% 5315 26056 7876 2566

Rotherham 12.3% 62.2% 17.3% 8.1% 71.0% 19.7% 9.3% 25143 126925 35311 16591

Ryedale 11.2% 65.1% 19.0% 4.7% 73.3% 21.4% 5.3% 4927 28701 8358 2088

Scarborough 11.2% 68.8% 14.6% 5.3% 77.5% 16.5% 6.0% 10133 62101 13205 4817

Selby 11.7% 61.0% 20.0% 7.2% 69.1% 22.7% 8.2% 7652 39854 13086 4702

Sheffield 14.2% 57.1% 17.7% 10.9% 66.6% 20.6% 12.8% 61851 249105 77218 47723

Wakefield 13.5% 57.6% 21.4% 7.5% 66.6% 24.7% 8.6% 35253 150533 55896 19484

York 12.5% 61.8% 18.3% 7.3% 70.7% 20.9% 8.4% 20317 100412 29746 11936



33/68

32

4. Limitations

SurveyThe data used for these estimates are based on self-reported drinking behaviour from theGeneral Lifestyle Survey and this may be prone to recall bias. Recent research suggests thatcurrent survey designs to measure alcohol consumption are likely to lead to underestimates in

the size of the population being affected by alcohol-related harms. When recallingconsumption, drinkers can ignore occasional (even routine) heavy drinking sessions,underestimate drink sizes and their alcohol content, and exclude holidays at home or abroad,special occasions or celebratory times of year when drinking may increase substantially. Evenwhen recollection is complete, accuracy relies on individuals neither exaggerating norunderestimating consumption.22 The actual physical effects of alcohol consumption can alsohamper recall, with excessive consumption being linked to blackouts and forgetting events ona night out.23 Taxation data provide an alternative source of information on consumption,detailing total quantities of alcohol purchased. Comparisons between taxation and surveydata show that survey data only account for 60% of alcohol purchased.24 However, eventaxation data do not provide a full estimation of total quantities consumed, as they excludeconsumption through homebrew, that imported from abroad (legally or illegally) and thatconsumed abroad. Surveys can be developed to more accurately capture consumption databut further work is needed to obtain a complete understanding using this methodology.25

The General Lifestyle Survey includes people in private households. A household is definedas a single person or a group of people who have the address as their only or main residenceand who either share one meal a day or share the living accommodation;26 and does notinclude people in institutions such as prisons or care homes nor those who are homeless.27The people excluded from participation may have a different pattern of alcohol consumption,and so the estimates reflect alcohol use among residents of private households only. Not allparticipants selected for inclusion participated in the survey, possibly introducing selectionbias, if those that agreed to take part in the survey differed systematically from those who didnot.

PopulationThe analysis uses mid-2007 population estimates provided by the Office for NationalStatistics; the current populations of each local authority may be slightly different.Furthermore, the mid-2007 populations are provisional data.

ModellingThe model uses data from a number of sources and any inaccuracies in these sources willaffect the modelled estimates. For example, alcohol-attributable hospital admissions areestimates based on research-derived attributable fractions for each health condition.28

22Off Measure: How we underestimate the amount we drink. Alcohol Concern. December, 2009.

23Teenage drinking, alcohol availability and pricing: A cross-sectional study of risk and protective factors for

alcohol-related harms in school children. Bellis et al., 2009.24

Off Measure: How we underestimate the amount we drink. Alcohol Concern. December, 2009.25

Improving accuracy in recording alcohol consumption: A survey in greater manchester. Morleo et al., 2011.26

General Lifestyle Survey 2008. Definitions and Terms. Office for National Statistics, 2010.27

Personal communication with the Surveys Team, Office for National Statistics. June, 2011.28

Alcohol-attributable fractions for England. Alcohol-attributable mortality and hospital admissions. Jones et al.,2008.


34/68

33

The modelling process involves making assumptions that include: 1) that the relationshipsidentified in the national General Lifestyle Survey between alcohol consumption and age, sexand ethnicity are the same at the local authority level; and 2) that the modelled estimate for aparticular local authority is the expected prevalence based on its population characteristicsand does not incorporate any local factors or initiatives that may have an impact on the trueprevalence of drinking behaviour. As with all modelling, extreme values are more difficult to

predict. The enhanced methodology employed by the North West Public Health Observatoryto calculate these updated estimates results in relatively wide confidence intervals. This isbecause the confidence intervals take into account the uncertainty in the estimates of themodel parameters, rather than solely the population size.

5. Quality Assurance

The estimates presented here have undergone rigorous quality assurance checks. In additionto being sub ject to the North West Public Health Observatorys internal quality assuranceprocesses, the main principles of the methodology and the application of the modelparameters to the population data have been externally checked by Yorkshire and Humber

Public Health Observatory as part of their inclusion in the 2011 Health Profiles.29

6. Conclusions

Directors of Public Health and others involved with the health of the population require localinformation on which to base commissioning decisions about resource allocation and tomeasure changes over time, particularly in relation to evaluating the effectiveness ofinterventions on reducing the proportion of the population drinking above governmentguidelines. Currently, local data on the extent of alcohol use and misuse are unavailable somonitoring the level of alcohol consumption within local populations remains a challenge.

These updated synthetic estimates of increasing risk and higher risk drinkers, together withnew estimates of abstainers and lower risk drinkers provide the best available nationalintelligence for commissioners to understand the current level of alcohol misuse in their localareas. The existence of large geographical variations demonstrates the need for local surveysand intelligence that can provide bespoke information. Local authorities should use theexperimental data presented here in combination with other local sources of information inorder to develop and direct their responses to alcohol harm.

According to the General Lifestyle Survey, the proportion of the population who abstain fromalcohol use has been rising steadily; about 7% of men abstained 1998 compared to about12% in 2009 while the comparable figures for women are 13% and 18% respectively.30 The

provision of estimates of levels of abstinence at local authority level is of particular importancein understanding the true nature and scale of the problem. Despite evidence to show that theyare already more likely to avoid venturing into towns at night because of alcohol-relatedbehaviour,31 abstainers are often ignored in the drinking debate. 32 Understanding the full

29http://www.apho.org.uk/default.aspx?QN=P_HEALTH_PROFILES


31Opinions on the impact of alcohol on individuals and communities: early summary findings from the North

West Big Drink Debate. Cook et al. 2008.32

The seven key messages of the alcohol industry. Addiction Info Switzerland et al., 2011.


35/68

34

range of drinking behaviours, from abstinence to higher risk is a critical element in planningactions to reduce consumption and avoid marginalisation of abstainers by policy makers.Local evaluations of drinking behaviour must, for example, be able to distinguish betweenchanges in levels of abstinence which simply reflect demographic changes in age andethnicity, and interventions which genuinely move higher and increasing risk drinkers into alower category of harm.

The estimates presented here should be used in context with other local and national alcoholdata, much of which is provided by the LAPE (Local Alcohol Profiles for England) tool athttp://www.nwph.net/alcohol/lape/. These estimates will be made available through that tooland additional data, including estimates of levels of dependent drinkers, will be added overcoming months.

7. AcknowledgementsThanks go to Mark Robinson, Simon Chambers and Sacha Wyke from the North West PublicHealth Observatory for the work they have conducted on quality assuring the estimates, toNeil Potter, also from the North West Public Health Observatory for preparing the maps, andto Michela Morleo and Elaine Steele of the Centre for Public Health for their comments on thereport. We would also like to thank the Depart of Health for commissioning this work and fortheir helpful comments on an earlier draft.
http://www.nwph.net/alcohol/lape/http://www.nwph.net/alcohol/lape/http://www.nwph.net/alcohol/lape/


36/68

35

Appendix 1: Methodological Summary

There are no direct, robust and consistent measures available for the number of people within local authoritieswho abstain, or who drink at lower risk, increasing risk or higher risk levels. Comparable and consistentmeasures of alcohol consumption are available from national lifestyle surveys that measure individual dinkinglevels and this type of survey can be used to derive synthetic estimates; here we use the General LifestyleSurvey (2008) as the primary data source.

The General Lifestyle Survey (until 2008 known as the General Household Survey) is a module of the IntegratedHousehold Survey and asks household respondents aged 16 and over across Great Britain a range of questionsincluding those related to alcohol use. Respondents are asked how often over the last year they have drunknormal strength beer, strong beer (6% or greater ABV

33), wine, spirits, fortified wines and alcopops; and how

much they usually drink on any one day. This information is combined to give an estimate o f the respondentsweekly alcohol consumption (averaged over a year) in units of alcohol. The method used for calculating usualweekly alcohol consumption is to multiply the number of units of each type drunk on a usual drinking day by thefrequency with which it was drunk using the factors shown below, and then to total across all drinks.

Drinking frequency Multiplying factor

Almost every day 7.0

5 or 6 days a week 5.5

3 or 4 days a week 3.5

Once or twice a week 1.5

Once or twice a month 0.375 (1.5 4)

Once every couple of months 0.115 (6 52)

Once or twice a year 0.029 (1.5 52)34

While the General Lifestyle Survey provides a measure of an individuals average weekly alcohol consumption,the data are typically only reported and considered robust at a regional (Government Office) geography. Onrequest, data can be provided at lower geographies, but because of sample sizes (total survey approximately

9,000 households per annum35

) local estimates based on survey data alone are not sufficiently robust.

In order to calculate synthetic estimates, data on adults aged 16 and over and resident in England wereextracted from the 2008 General Lifestyle Survey (N=14,036). Records which contained missing data for thevariables later used in the analysis (age, sex, ethnicity and average weekly alcohol consumption) were removedleaving a final data set containing 12,470 individual records. A variable drinking category was created for eachindividual based upon their value for average weekly alcohol consumption (see Table 1 for definitions): abstainer(N=1,766), lower risk (N=7,794), increasing risk (N=2,187) and higher risk (N=723).

The General Lifestyle Survey dataset contained a respondents Lower Super Output Area36

of residence and thiswas used to identify their local authority of residence. Local authority level data were appended to eachindividual record in the dataset. These local authority level measures are detailed in Table 11. All continuousvariables were then categorised and ethnicity was re-categorised into a smaller number of categories; details ofthe categories used in the modelling process are reported in Table 12.

33Alcohol by volume: standard measure of how much alcohol is contained within an alcoholic drink.


35Summary Quality Report for General Lifestyle Survey (GLF) Releases. Office for National Statistics, 2009.

36These are small geographical areas designed for the collection of data for small area statistics.


37/68

36

Table 11: Local authority level measures included in the modelling

Variable Year Description Source

Index ofMultipleDeprivation

2007 A composite measure of deprivation for small areas.The overall score is a combination of sevendomains, each including a number of indicators.

Department for Communitiesand Local Government.

The income

domain of theIndex ofMultipleDeprivation

2007 One of the seven domains included within the Index

of Multiple Deprivation.

Department for Communities

and Local Government.

Alcohol-specifichospitaladmissions

2008/09 Admissions for conditions that are wholly related toalcohol (for example, alcoholic liver disease oralcohol overdose).

Calculated using HospitalEpisode Statistics.

Alcohol-attributablehospital

admissions

2008/09 Admissions for alcohol-specific conditions plusconditions that are caused by alcohol in some, butnot all, cases (for example, stomach cancer and

unintentional injury).*

Calculated using HospitalEpisode Statistics.

Alcohol-specificmortality

2006 to2008

Mortality from conditions that are wholly related toalcohol (for example, alcoholic liver disease oralcohol overdose).

Calculated using mortalitydata.

Alcohol-attributablemortality

2008 Mortality from alcohol-specific conditions plusconditions that are caused by alcohol in some, butnot all, cases (for example, stomach cancer andunintentional injury).*

Calculated using mortalitydata.

Alcohol-attributable

recordedcrime

2009/10 Aggregation of six alcohol-related offences: violenceagainst the person, sexual offences, robbery,

burglary dwelling, theft of a motor vehicle and theftfrom a vehicle.**

Home Office recorded crimestatistics.

StrategicHealthAuthority

2004 2004 configuration; 28 geographic areas. Office for National Statistics.

*A list of alcohol-attributable conditions with their ICD-10 codes and the method for calculating alcohol-attributable conditions can be found at:http://www.nwph.net/nwpho/publications/AlcoholAttributableFractions.pdf**Attributable fractions for alcohol-related crimes (i.e. the percentage of people arrested for a particular type ofcrime who test positive for alcohol during a urine test) were taken from the Home Office New English and WelshArrestee Drug Abuse Monitoring System (NEW-ADAM) arrestee survey (1999-2001). Data were taken from 16police stations in England and Wales; only offences with sample sizes of more than 50 arrestees have beenincluded. Intoxicated arrestees were not interviewed, which suggests that some figures are likely to beunderestimates and explains why drunkenness offences are not included. These attributable fractions for eachcrime were then applied to recorded crime data.
http://www.nwph.net/nwpho/publications/AlcoholAttributableFractions.pdfhttp://www.nwph.net/nwpho/publications/AlcoholAttributableFractions.pdfhttp://www.nwph.net/nwpho/publications/AlcoholAttributableFractions.pdfhttp://www.nwph.net/nwpho/publications/AlcoholAttributableFractions.pdf


38/68

37

Table 12: Categorisation of the variables used in the modelling process

Variable Categorisation

Individual level variables

Age 16-19, 20-24, 25-29, 30-34..65-69, 70-74 and 75 and over.

Ethnic group White, Asian, Black/Other.

Local authority level variables

Index of Multiple Deprivation Less than 11.0 (quintile 1), 11.0 to 14.9 (quintile 2), 15.0 to 20.4 (quintile 3),20.5 to 26.5 (quintile 4), 26.6 or above (quintile 5).

Income domain of the Indexof Multiple Deprivation:

Less than 0.087 (quintile 1), 0.087 to 0.110 (quintile 2), 0.111 to 0.0141(quintile 3), 0.142 to 0.182 (quintile 4), 0.183 or above (quintile 5).

Alcohol-specific mortality Less than 5.5 (quintile 1), 5.5 to 7.12 (quintile 2), 7.13 to 9.21 (quintile 3),9.22 to 11.9 (quintile 4), 12.0 and above (quintile 5).

Alcohol-attributable mortality Less than 19.74 (quintile 1), 19.74 to 22.37 (quintile 2), 22.38 to 25.74

(quintile 3), 25.75 to 29.74 (quintile 4), 29.75 and above (quintile 5).

Alcohol-specific hospitaladmissions


Alcohol-attributable hospitaladmissions


Alcohol-attributable recordedcrime

Less than 4.71 (quintile 1), 4.71 to 6.03 (quintile 2), 6.04 to 7.54 (quintile 3),7.55 to 9.47 (quintile 4), 9.48 or above (quintile 5).

Forward, stepwise, multinomial logistic regression modelling was used to identify which of the variables were

significantly associated with drinking category, using SPSS statistical software (version 16). Another 20 modelswere created using forward, stepwise, multinomial logistic regression, each based upon a randomly selectedsample of records containing 80% of the total data in order to test robustness of the contribution of differentvariables. The income domain was highly correlated with the Index of Multiple Deprivation, so the latter wasincluded in the final model. Alcohol-attributable and alcohol-specific hospital admissions were highly correlated,as were alcohol-attributable and alcohol-specific mortality; alcohol-specific mortality and alcohol-attributablehospital admissions were chosen for inclusion in the final model because they were significantly associated withdrinking category in a greater proportion of the exploratory models. The final model therefore contained sevenvariables: age, sex, ethnicity, Strategic Health Authority, Index of Multiple Deprivation, alcohol-specific mortalityand alcohol-attributable hospital admissions. Using abstainers as the baseline category, the final model wasused to predict the probability of being an abstainer, a lower risk drinker, an increasing risk drinker and a higherrisk drinker according to age, sex, ethnicity and so on (see Tables 13 to 15 for the model parameters).

Mid-2007 population estimates for local authorities were obtained through a bespoke request to the Office forNational Statistics and were grouped into combinations by age group/sex/ethnicity (for example, of females,aged 20 to 24 who were Asian). An estimate of the 16 to 19 year olds population was calculated by assumingthe population for the 15 to 19 quinary age band was evenly spread across the individual age bands. Thesepopulations were multiplied by the predicted probabilities obtained from the multinomial logistic regression modelfor abstainers, lower risk, increasing risk and higher risk drinkers in the respective age/sex/ethnic group for eachlocal authority. Overall prevalence estimates for each drinking category were then obtained by summing thepredicted number of abstainers, lower risk, increasing risk and higher risk drinkers within all sex/age/ethnicitygroups in each local authority. Using Monte Carlo methods, 95% confidence intervals were generated for eachlocal authority (see Appendices 2 and 3 for further details of the modelling and calculation of confidence intervalsrespectively).


39/68

38

Appendix 2: Multinomial Regression Model

The overall process of producing local authority level prevalence estimates for abstainers, lower risk, increasingrisk and higher risk drinking involved a number of steps:

1. The probability of being an abstainer, lower risk, increasing risk or higher risk drinker was modelledusing multinomial logistic regression as a function of variables measured at the individual (age, sex,ethnicity) and area (index of multiple deprivation, alcohol-attributable hospital admissions and alcohol-specific mortality, Strategic Health Authority) level.

2. The model was used to generate estimated probabilities of abstainers, lower risk, increasing risk, andhigher risk drinking by age group, sex and ethnicity for all English local authorities.

3. Age, sex and ethnicity specific probabilities were applied to the age, sex and ethnicity specific populationestimates for each local authority to provide an estimate of the overall number, and prevalence (as apercentage) of abstainers, lower risk, increasing risk and higher risk drinkers in each area.

Two regression models were considered for modelling the data, namely, ordinal and multinomial logisticregression. The latter was chosen as there was evidence that the assumption of proportionality for the ordinalregression model did not hold. Only the multinomial regression results are reported for which predictedprobabilities were obtained for the selected covariates in the multinomial logistic regression model:

for s 1

Wheres are the states:

1. Abstainers2. Lower risk drinkers3. Increasing risk drinkers4. Higher risk drinkers

is the vector of attributes of the ith local authority

is the vector of coefficients for state s (tables 13 to 15)

t are the vector of coefficients for states 2 to 4


40/68

39

Table 13:Estimated model parameters for lower risk drinking with respect to the base category (abstainers)

Lower risk drinking B Exp(B)

95% ConfidenceInterval for Exp(B)

LowerBound

UpperBound

Intercept -1.982

Male .536 1.709 1.524 1.917

Female 0 . . .

White 2.452 11.611 9.058 14.884

Black/other .793 2.209 1.621 3.011

Asian 0 . . .

Norfolk .152 1.164 .791 1.712

Bedfordshire and Hertfordshire .031 1.031 .664 1.602

Essex .334 1.397 .873 2.234

North West London -.038 .962 .591 1.567

North Central London -.287 .750 .457 1.233

North East London -.526 .591 .356 .979

South East London .024 1.024 .635 1.652

South West London -.162 .850 .547 1.323

Northumberland -.310 .734 .477 1.129

County Durham and Tees Valley .062 1.064 .643 1.760

North and East Yorkshire andNorthern Lincolnshire

.036 1.037 .679 1.584

West Yorkshire .088 1.092 .721 1.655

Cumbria and Lancashire .136 1.146 .759 1.729

Greater Manchester .196 1.216 .815 1.815

Cheshire & Merseyside -.023 .978 .660 1.448

Thames Valley -.186 .830 .552 1.249

Hampshire and Isle of Wight .375 1.455 .943 2.247

Kent and Medway .117 1.124 .723 1.748

Surrey and Sussex .316 1.372 .906 2.079

Avon .358 1.431 .951 2.153

South West Peninsula .323 1.381 .876 2.176

Dorset and Somerset -.149 .861 .560 1.325

South Yorkshire .143 1.154 .742 1.795

Trent .158 1.171 .809 1.697

Leicestershire .114 1.120 .728 1.724

Shropshire and Staffordshire .342 1.408 .899 2.204

Birmingham and the Black Country .054 1.055 .709 1.572

West Midlands South 0 . . .

16-19 yrs .433 1.542 1.177 2.020

20-24 yrs 1.271 3.563 2.557 4.964

25-29 yrs .979 2.662 2.026 3.496

30-34 yrs 1.052 2.865 2.217 3.702

35-39 yrs 1.260 3.524 2.733 4.544

40-44 yrs 1.191 3.289 2.572 4.207

45-49 yrs .999 2.715 2.132 3.456

50-54 yrs 1.071 2.918 2.251 3.783


41/68

40

Lower risk drinking B Exp(B)95% Confidence

Interval for Exp(B)

55-59 yrs .850 2.339 1.844 2.966

60-64 yrs .697 2.007 1.608 2.504

65-69 yrs .450 1.568 1.251 1.966

70-74 yrs .455 1.576 1.246 1.992

75 yrs and over 0 . . .

IMD q1 .185 1.203 .866 1.672

IMD q2 .216 1.241 .934 1.649

IMD q3 .185 1.204 .938 1.545

IMD q4 .150 1.162 .950 1.421

IMD q5 0 . . .

AS Mortality q1 -.117 .890 .644 1.228

AS Mortality q2 .110 1.117 .835 1.494

AS Mortality q3 -.080 .923 .725 1.174

AS Mortality q4 .028 1.028 .841 1.258

AS Mortality q5 0 . . .

AA Hosp Admissions q1 .327 1.386 .991 1.940AA Hosp Admissions q2 .098 1.103 .818 1.488

AA Hosp Admissions q3 .150 1.162 .885 1.526


AA Hosp Admissions q5 0 . . .

IMD: Index of Multiple Deprivation. AS: alcohol-specific. AA Hosp: alcohol-attributable hospital. q: quintile.


42/68

41

Table 14:Estimated model parameters for increasing risk drinking with respect to the base category (abstainers)

Increasing risk drinking B Exp(B)


LowerBound

UpperBound

Intercept -5.124

Male .935 2.546 2.219 2.921

Female 0 . . .

White 3.371 29.107 17.369 48.776

Black/other .631 1.879 .946 3.734

Asian 0 . . .

Norfolk -.193 .824 .505 1.345

Bedfordshire and Hertfordshire .108 1.115 .652 1.905

Essex .122 1.130 .634 2.013

North West London .005 1.005 .520 1.943

North Central London .068 1.070 .555 2.061

North East London -.045 .956 .488 1.875

South East London .485 1.623 .898 2.935

South West London .239 1.270 .733 2.202

Northumberland -.045 .956 .559 1.635


North and East Yorkshire and NorthernLincolnshire

.094 1.098 .658 1.832

West Yorkshire .627 1.871 1.132 3.094

Cumbria and Lancashire .513 1.670 1.019 2.737


Cheshire & Merseyside .005 1.005 .617 1.637

Thames Valley -.206 .813 .493 1.342


Kent and Medway -.190 .827 .474 1.443

Surrey and Sussex .640 1.897 1.161 3.101

Avon .420 1.522 .934 2.481

South West Peninsula .910 2.485 1.448 4.264


South Yorkshire .452 1.572 .915 2.700

Trent .211 1.235 .782 1.953

Leicestershire .326 1.385 .819 2.343


Birmingham and the Black Country -.129 .879 .522 1.481


16-19 yrs 1.031 2.804 1.949 4.035

20-24 yrs 2.100 8.169 5.448 12.248

25-29 yrs 1.717 5.569 3.913 7.925

30-34 yrs 1.604 4.971 3.536 6.990

35-39 yrs 2.031 7.625 5.511 10.551

40-44 yrs 1.987 7.293 5.323 9.994

45-49 yrs 1.810 6.109 4.470 8.347

50-54 yrs 1.946 6.998 5.044 9.709


43/68

42

Increasing risk drinking B Exp(B)95% Confidence

Interval for Exp(B)

55-59 yrs 1.703 5.492 4.036 7.473

60-64 yrs 1.204 3.333 2.464 4.510

65-69 yrs 1.032 2.805 2.057 3.827

70-74 yrs .719 2.053 1.471 2.864


IMD q1 .528 1.696 1.138 2.527

IMD q2 .511 1.667 1.180 2.357

IMD q3 .525 1.690 1.247 2.292

IMD q4 .097 1.102 .861 1.412

IMD q5 0 . . .

AS Mortality q1 -.207 .813 .550 1.202

AS Mortality q2 -.038 .963 .676 1.371

AS Mortality q3 -.253 .777 .580 1.041

AS Mortality q4 -.021 .979 .769 1.247


AA Hosp Admissions q1 .220 1.247 .827 1.880AA Hosp Admissions q2 -.012 .988 .681 1.434


AA Hosp Admissions q4 -.088 .916 .696 1.205

AA Hosp Admissions q5 0 . . .

IMD: Index of Multiple Deprivation. AS: alcohol-specific. AA Hosp: alcohol-attributable hospital. q: quintile.


44/68

43

Table 15:Estimated model parameters for higher risk drinking with respect to the base category (abstainers)

Higher risk drinking B Exp(B)


LowerBound

UpperBound

Intercept -6.888

Male 1.043 2.837 2.361 3.409

Female 0 . . .

White 3.176 23.959 11.498 49.923

Black/other .578 1.782 .667 4.765

Asian 0 . . .

Norfolk -.247 .781 .385 1.585

Bedfordshire and Hertfordshire -.130 .878 .401 1.921

Essex .403 1.496 .682 3.283

North West London 1.201 3.323 1.489 7.418

North Central London .113 1.120 .453 2.772

North East London .646 1.907 .781 4.661

South East London .819 2.268 1.018 5.051

South West London .664 1.942 .917 4.112

Northumberland .363 1.438 .692 2.986


North and East Yorkshire and NorthernLincolnshire

.428 1.535 .784 3.005

West Yorkshire .944 2.571 1.322 5.001

Cumbria and Lancashire .808 2.243 1.155 4.357


Cheshire & Merseyside .229 1.257 .637 2.480

Thames Valley .042 1.043 .529 2.056


Kent and Medway .151 1.163 .537 2.518

Surrey and Sussex .701 2.017 1.040 3.910

Avon .080 1.083 .542 2.165

South West Peninsula .894 2.444 1.151 5.189


South Yorkshire 1.009 2.742 1.359 5.533

Trent .599 1.820 .975 3.397

Leicestershire .159 1.173 .556 2.476


Birmingham and the Black Country -.324 .723 .328 1.595


16-19 yrs 1.972 7.188 4.001 12.914

20-24 yrs 2.861 17.485 9.439 32.390

25-29 yrs 2.278 9.761 5.411 17.608

30-34 yrs 2.206 9.083 5.097 16.185

35-39 yrs 2.668 14.408 8.335 24.905

40-44 yrs 2.661 14.311 8.379 24.444

45-49 yrs 2.752 15.670 9.275 26.475

50-54 yrs 2.588 13.307 7.692 23.023


45/68

44

Higher risk drinking B Exp(B)95% Confidence

Interval for Exp(B)

55-59 yrs 2.205 9.071 5.279 15.587

60-64 yrs 2.210 9.115 5.424 15.317

65-69 yrs 1.745 5.725 3.314 9.891

70-74 yrs 1.321 3.747 2.081 6.748


IMD q1 .756 2.129 1.244 3.643

IMD q2 .363 1.438 .903 2.290

IMD q3 .299 1.349 .895 2.033

IMD q4 .154 1.166 .840 1.618

IMD q5 0 . . .

AS Mortality q1 -.733 .480 .283 .816

AS Mortality q2 -.389 .678 .422 1.087

AS Mortality q3 -.660 .517 .347 .769

AS Mortality q4 -.372 .690 .499 .954


AA Hosp Admissions q

Alcohol Estimates 2011

Documents

Transcript of Alcohol Estimates 2011